RealHD: A High-Quality Dataset for Robust Detection of State-of-the-Art AI-Generated Images

TL;DR

RealHD introduces a high-quality, large-scale dataset with over $730{,}000$ images spanning real and AI-generated content across T2I, inpainting, refinement, and face swapping, accompanied by rich annotations and a comprehensive prompt-generation pipeline. It demonstrates that detectors trained on RealHD generalize better across modalities and domains, addressing prior generalization gaps. The authors also propose a lightweight image noise entropy (NE) detector based on non-local means residuals and block-wise Shannon entropy, achieving competitive performance and improved robustness to compression. Together, RealHD and NE establish a strong baseline for robust detection of AI-generated images and provide a practical benchmark for future research and cross-domain evaluation.

Abstract

The rapid advancement of generative AI has raised concerns about the authenticity of digital images, as highly realistic fake images can now be generated at low cost, potentially increasing societal risks. In response, several datasets have been established to train detection models aimed at distinguishing AI-generated images from real ones. However, existing datasets suffer from limited generalization, low image quality, overly simple prompts, and insufficient image diversity. To address these limitations, we propose a high-quality, large-scale dataset comprising over 730,000 images across multiple categories, including both real and AI-generated images. The generated images are synthesized via state-of-the-art methods, including text-to-image generation (guided by over 10,000 carefully designed prompts), image inpainting, image refinement, and face swapping. Each generated image is annotated with its generation method and category. Inpainting images further include binary masks to indicate inpainted regions, providing rich metadata for analysis. Compared to existing datasets, detection models trained on our dataset demonstrate superior generalization capabilities. Our dataset not only serves as a strong benchmark for evaluating detection methods but also contributes to advancing the robustness of AI-generated image detection techniques. Building upon this, we propose a lightweight detection method based on image noise entropy, which transforms the original image into an entropy tensor of Non-Local Means (NLM) noise before classification. Extensive experiments demonstrate that models trained on our dataset achieve strong generalization, and our method delivers competitive performance, establishing a solid baseline for future research. The dataset and source code are publicly available at https://real-hd.github.io.

Figures (2)

• Figure 1: Overview of the prompt construction pipeline. The process consists of three key stages: (1) constructing a corpus repository containing 15 sub-corpora categorized by theme, (2) generating short prompts by instantiating predefined templates with expressions sampled from the sub-corpora, and (3) refining and enriching the prompts using large language models (LLMs) to improve fluency, semantic richness, and contextual diversity.
• Figure 2: Comparison of noise entropy distributions between real and generated images in the RealHD and GenImage datasets (based on 1,000 sampled images). These differences motivate a lightweight detection method based on image noise entropy, which improves model generalizability.